Guaranteed non-asymptotic confidence regions in system identification

In this paper we consider the problem of constructing confidence regions for the parameters of identified models of dynamical systems. Taking a major departure from the previous literature on the subject, we introduce a new approach called ‘Leave-out Sign-dominant Correlation Regions’ (LSCR) which d...

Full description

Saved in:
Bibliographic Details
Published inAutomatica (Oxford) Vol. 41; no. 10; pp. 1751 - 1764
Main Authors Campi, M.C., Weyer, E.
Format Journal Article
LanguageEnglish
Published Oxford Elsevier Ltd 01.10.2005
Elsevier
Subjects
Applied sciences
Computer science; control theory; systems
Confidence sets
Control theory. Systems
Exact sciences and technology
Finite sample results
General linear models
Modelling and identification
System identification
Uncertainty evaluation
Uncertainty evaluation
Confidence sets
General linear models
System identification
Finite sample results
Correlation
Uncertain system
Probabilistic approach
Confidence
Asymptotic approximation
Dynamical system
Modeling
Online AccessGet full text

Cover

More Information
Summary:In this paper we consider the problem of constructing confidence regions for the parameters of identified models of dynamical systems. Taking a major departure from the previous literature on the subject, we introduce a new approach called ‘Leave-out Sign-dominant Correlation Regions’ (LSCR) which delivers confidence regions with guaranteed probability. All results hold rigorously true for any finite number of data points and no asymptotic theory is involved. Moreover, prior knowledge on the noise affecting the data is reduced to a minimum. The approach is illustrated on several simulation examples, showing that it delivers practically useful confidence sets with guaranteed probabilities.
ISSN:0005-1098
1873-2836
DOI:10.1016/j.automatica.2005.05.005